HP Forums
DM42 News - Printable Version

+- HP Forums (https://www.hpmuseum.org/forum)
+-- Forum: Not HP Calculators (/forum-7.html)
+--- Forum: Not quite HP Calculators - but related (/forum-8.html)
+--- Thread: DM42 News (/thread-7457.html)

Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22


RE: DM42 News - larthurl - 12-19-2017 09:03 PM

(12-19-2017 10:25 AM)Paul Dale Wrote:  He who must not be named asked me to post this.

Pauli

(12-19-2017 08:43 AM)emersone12 Wrote:  If Swiss Micros have only produced an initial batch of 300 production models starting from serial number 100 & then numbering sequentially onwards from SN 100 then theoretically the highest serial number should logically be SN 400. So how can there be SN 429?

Hi emersone12,

Did you ever hear about scrap rate?

Mine is #310.


RE: DM42 News - rprosperi - 12-19-2017 09:10 PM

(12-19-2017 07:58 PM)Logan Wrote:  I don't know if this is a bug or not, but on the original HP 42s when using the Alpha entry, you could hold down the shift the entire time and select lowercase characters without letting go. That doesn't seem to be the case with DM42 (or Free42 iOS), where it requires shift for each additional character.

They are different, likely due to the varying nature of keyboards that Free42 works with (logical, if you think about it).

But on the DM42, once in Alpha mode, use [Shift][Down] until you see the [a] indicator, and you are in lower case mode and then no need to shift at all for as many lc chars using the full keyboard as you want. [Shift][Up] to go back to [A] (upper case) and use the full keyboard as well, [Shift][Up] again to go back to [ ] mode, when it behaves just like a real 42S (except of course the hold-shift thing). Repeated [Shift]-[Up] or -[Down] will wrap around the 3 states.


RE: DM42 News - Logan - 12-19-2017 09:15 PM

(12-19-2017 09:10 PM)rprosperi Wrote:  They are different, likely due to the varying nature of keyboards that Free42 works with (logical, if you think about it).

But on the DM42, once in Alpha mode, use [Shift][Down] until you see the [a] indicator, and you are in lower case mode and then no need to shift at all for as many lc chars using the full keyboard as you want. [Shift][Up] to go back to [A] (upper case) and use the full keyboard as well, [Shift][Up] again to go back to [ ] mode, when it behaves just like a real 42S (except of course the hold-shift thing). Repeated [Shift]-[Up] or -[Down] will wrap around the 3 states.

Thanks Bob, I had a question earlier in this thread about how to use the full keyboard characters so I can do that...I just sometimes like to use the old-style alpha menu for short variables and it is rather nice to be able to hold the shift key down.

I figured it might have originally been this way because some devices might not accept simultaneous inputs, but thought I'd mention it in the case of DM42 anyway. It may have been discussed at some point on the Swiss Micros forums, I know I didn't read everything over there.


RE: DM42 News - rprosperi - 12-19-2017 09:31 PM

(12-19-2017 09:15 PM)Logan Wrote:  Thanks Bob, I had a question earlier in this thread about how to use the full keyboard characters so I can do that...I just sometimes like to use the old-style alpha menu for short variables and it is rather nice to be able to hold the shift key down.

I figured it might have originally been this way because some devices might not accept simultaneous inputs, but thought I'd mention it in the case of DM42 anyway. It may have been discussed at some point on the Swiss Micros forums, I know I didn't read everything over there.

Sure, no problem. I should have said (logical, if you think about it, which I hadn't either until this question).

I too often find I am using the 'classic style' Alpha menu without thinking about it, I guess from years of using a real 42S. But in those years, I don't think I discovered the trick of holding down shift like this. I rarely used lower case Alpha except for certain output labels, specifically because it was a pain to use.


RE: DM42 News - Anders - 12-20-2017 12:29 AM

Hooray! Got mine today #215 :-)


RE: DM42 News - Benjer - 12-20-2017 01:02 AM

Arrived here in Oregon.

#368.


RE: DM42 News - 4ster - 12-20-2017 01:32 AM

# 401 arrived in Washington state today.

I think some thank you's are in order. First, to Thomas Okken for developing Free42 over the years. Next, to Swiss Micros for the vision to bring the ideal platform for Free42 into being. I can't begin to understand how much work it was to create the hardware as it is designed. And lastly to those of you who beta tested this calculator. THANK YOU all!

This is the first "new model" calculator I've purchased in nearly 30 years. (I kind of stopped buying calculators after the HP 41. I did purchase a DM-41 a while ago, but that doesn't count since it is a mini-me of the 41.)

I think it was worth the wait.


RE: DM42 News - Thomas Okken - 12-20-2017 01:38 AM

(12-19-2017 08:21 PM)salvomic Wrote:  
(12-19-2017 07:09 PM)Bunuel66 Wrote:  ...
I have checked the behavior of Free42 on Windows and Android. The behavior is similar: takes way longer if the domain is exactly 0,Pi. The problem appears only on the decimal version. The binary just works fine. On Windows the integration
ends up slowly but it ends. ...

(12-19-2017 07:30 PM)Thomas Okken Wrote:  Bug, certainly. Probably in the termination condition.
I'll have to look into that.

Free42 in iOS 11 (iPhone), after few seconds (about 10-15) with acc. 1E-20 I get -5.731411E-34 (and not 0)

Hmmm. I guess it's probably not so much a bug as a difficult case. Error estimates get tricky when the exact result is zero. Still, something to look into.


RE: DM42 News - Thomas Okken - 12-20-2017 01:48 AM

(12-19-2017 07:58 PM)Logan Wrote:  I don't know if this is a bug or not, but on the original HP 42s when using the Alpha entry, you could hold down the shift the entire time and select lowercase characters without letting go. That doesn't seem to be the case with DM42 (or Free42 iOS), where it requires shift for each additional character.

As Bob suggested, that's because the touchscreens that Free42 was originally designed to work with were all single-touch, so it was impossible to press two or more keys at once, and so there was never a reason for Free42 to handle that scenario.

Now that we have multi-touch screens and the DM42, it would make more sense to support true HP-42S Shift behavior... but it hasn't been done yet, obviously, and I'm not making any promises if or when it will get done.

(12-19-2017 09:31 PM)rprosperi Wrote:  I too often find I am using the 'classic style' Alpha menu without thinking about it, I guess from years of using a real 42S. But in those years, I don't think I discovered the trick of holding down shift like this. I rarely used lower case Alpha except for certain output labels, specifically because it was a pain to use.

I think it's mentioned in the manual somewhere. I've known about this Shift behavior forever, but it's not something I discovered on my own. I must have either read it in the 42S manual, or else I read about this type of behavior on the HP-48G, and then got curious and started wondering if that might work on the 42S as well.

UPDATE: Yep, it's in the manual: page 37, under ALPHA Typing Tips.


RE: DM42 News - rprosperi - 12-20-2017 02:26 AM

(12-20-2017 01:48 AM)Thomas Okken Wrote:  UPDATE: Yep, it's in the manual: page 37, under ALPHA Typing Tips.

LOL, I read this post, then went to look it up, and found it. By the time I clicked "Quote" to confirm it, your update appeared.

As many times as I've read the 42S manual, I still find stuff (like this) that I've clearly read, yet somehow missed.


RE: DM42 News - Werner - 12-20-2017 07:11 AM

#345 is in Belgium!
What a splendid piece of work it is!

Cheers, Werner


RE: DM42 News - Michael de Estrada - 12-20-2017 03:12 PM

I just realized that my DM42 is missing the programming (PGM) hole at the back of the case. There is no PGM label either. Has this occured on any of the other units ? I’ve also posted this question in the SM forum.


RE: DM42 News - grsbanks - 12-20-2017 03:14 PM

(12-20-2017 03:12 PM)Michael de Estrada Wrote:  I just realized that my DM42 is missing the programming (PGM) hole at the back of the case. There is no PGM label either. Has this occured on any of the other units ? I’ve also posted this question in the SM forum.

...where I replied that it's not needed Smile

As far as I know, none of the production units have it and I never needed to use it on my beta model either.


RE: DM42 News - Jeroen Van Nieuwenhove - 12-23-2017 01:39 AM

#101 arrived in Belgium yesterday. *Very* nice build quality, even better than the DM 41L I bought some time ago from swissmicros. Really excellent display. Keys are quite OK, my only quibble being that the Enter key must be pressed on the right side to register reliably. The pouch is a bit too snug but otherwise very nice.
Being a regular user of Thomas' Free42 on my PC and smartphone, this standalone hardware implementation is a real treat to use. My thanks to Michael, Thomas and of course all beta testers for this terrific reincarnation of one of HP's best calcs.


RE: DM42 News - markhen - 12-25-2017 08:20 AM

#331 arrived in Los Angeles late this week. Absolutely gorgeous calculator - I have almost every HP model and this really impresses. Thank you SM and TO!!!


RE: DM42 News - badaze - 12-27-2017 12:07 AM

Received mine yesterday. SN #142.


RE: DM42 News - pascal_meheut - 12-27-2017 01:49 AM

(12-27-2017 12:07 AM)badaze Wrote:  Received mine yesterday. SN #142.

Did too. SN #462. As described here: pouch, keyboard and so on...
Nice calculator.


RE: DM42 News - Michael de Estrada - 12-27-2017 02:51 AM

(12-27-2017 01:49 AM)pascal_meheut Wrote:  
(12-27-2017 12:07 AM)badaze Wrote:  Received mine yesterday. SN #142.

Did too. SN #462. As described here: pouch, keyboard and so on...
Nice calculator.

I thought they only made 300 and the production serial numbers began at 101, so how did you end up with SN #462 ?


RE: DM42 News - Thomas Okken - 12-27-2017 03:11 AM

INTEG(0, pi, cos(x)), continued --
(Apologies for the messy post. The forum's quoting logic is acting up.)

(12-20-2017 01:38 AM)Thomas Okken Wrote:  
(12-19-2017 08:21 PM)salvomic Wrote:  Free42 in iOS 11 (iPhone), after few seconds (about 10-15) with acc. 1E-20 I get -5.731411E-34 (and not 0)

Hmmm. I guess it's probably not so much a bug as a difficult case. Error estimates get tricky when the exact result is zero. Still, something to look into.

I took another look at the code, and it's fine, but it doesn't handle cases like this. The termination condition is

| current_estimate - previous_estimate | <= ACC * | current_estimate |

that is, ACC is a relative error bound, and if the estimates keep getting smaller in an exponential fashion, the termination condition may take a long time to be met, especially if ACC = 0, which is the default (ACC unset does the same as ACC = 0). The iteration will terminate eventually: there is a hard-coded limit of 10 steps, which means 524,288 evaluations of the integrand.

Like I said, the error estimates get tricky when the exact result is zero. It makes me wonder how the real HP-42S decides to call it quits with this problem after only three iterations. Is it using a hard-coded absolute error bound, in addition to ACC?


RE: DM42 News - pascal_meheut - 12-27-2017 09:41 AM

(12-27-2017 02:51 AM)Michael de Estrada Wrote:  I thought they only made 300 and the production serial numbers began at 101, so how did you end up with SN #462 ?

No idea. I ordered it the 20th of December, received it the 26th under firmware 3.0.